Two crucial factors have urged IR scholars to account for state behavior in cyberspace: the increasing volume of cyber attacks and the rising scale of damage induced by these attacks. These have led to a series of initiatives appealing to the grand theories of IR to scrutinize state behavior in depth, particularly regarding the securitization and militarization of cyberspace. Following a similar pattern, this work intends to contribute to IR literature by presenting an analysis of state behavior in cyberspace by relying on quantitative methods. Two regression analyses, one using a linear algorithm and the other using a negative binomial algorithm, were conducted. The analyses indicate that as states build more cyber security capacity, they tend to engage in more disruptive actions against other states in cyberspace. Both regression models were subsequently trained using artificial neural networks (ANNs) for robustness check, which enhanced the results computed by the traditional models.